*** jamesmcarthur has joined #openstack-tc | 00:02 | |
fungi | dims: continuing from the earlier discussion, https://techcrunch.com/2019/02/27/open-source-communities-fight-over-telco-market/ has some really disappointing quotes in it | 00:11 |
---|---|---|
*** mriedem has quit IRC | 00:13 | |
*** jamesmcarthur has quit IRC | 00:17 | |
*** jamesmcarthur has joined #openstack-tc | 00:18 | |
*** mriedem has joined #openstack-tc | 00:19 | |
*** jamesmcarthur has quit IRC | 00:22 | |
*** lbragstad has quit IRC | 00:32 | |
*** dangtrinhnt has quit IRC | 00:40 | |
mnaser | fungi: yep, really disappointing | 00:41 |
fungi | i don't get why this has to be a competition instead of a collaboration | 00:43 |
zaneb | kudos to Boris for hitting the nail on the head though | 00:54 |
fungi | yep, thrilled to see him staying on-message ;) | 01:06 |
dtroyer | Hmmm… press + need for clicks == drama? Do I smell a reality show in here somewhere? | 01:14 |
* dtroyer sneaks out the back door | 01:14 | |
*** dangtrinhnt has joined #openstack-tc | 01:17 | |
*** bsilverman has joined #openstack-tc | 02:37 | |
*** mriedem has quit IRC | 03:32 | |
dims | fungi : Ack. that's very sad and disappointing piece | 03:46 |
dims | dtroyer : right, dunno how much of it was twisted to fit a narrative | 03:47 |
dims | +1 to Boris! | 03:47 |
*** spsurya has joined #openstack-tc | 04:14 | |
*** bsilverman has quit IRC | 05:38 | |
*** Luzi has joined #openstack-tc | 06:50 | |
*** jpich has joined #openstack-tc | 08:00 | |
*** tosky has joined #openstack-tc | 08:39 | |
*** jaosorior has joined #openstack-tc | 09:06 | |
*** e0ne has joined #openstack-tc | 09:32 | |
*** dtantsur|afk is now known as dtantsur | 09:49 | |
*** cdent has joined #openstack-tc | 11:00 | |
* ttx declares IRC backlog bankruptcy after a successful week off IRC -- reping me if there was anything urgent or important that was not replicated in email form | 11:04 | |
*** e0ne has quit IRC | 11:30 | |
*** e0ne has joined #openstack-tc | 11:36 | |
jroll | not a good look for the CNCF, for well-informed people. unfortunately that isn't the target for this article | 11:45 |
*** jpich has quit IRC | 11:47 | |
bauzas | hopefully it's just a vision from the CNCF only said by this article | 12:01 |
bauzas | so thanks to Boris and other folks, the article explains this vision is wrong | 12:01 |
cdent | the quoted rep from the cncf sounds rather combative | 12:01 |
bauzas | now, I'm a bit afraid of seeing this vision not only there but rather done with others | 12:02 |
bauzas | now, is it only an individual vision or the CNCF ? | 12:03 |
bauzas | I hope for the former | 12:03 |
ttx | well, Dan Kohn is the Executive Director of the CNCF, so it's hard to separate him from it | 12:05 |
cdent | I have to admit from a generate press standpoint, being combative is probably a useful tactic | 12:05 |
cdent | but it's not at all in the open source spirit | 12:06 |
cdent | hell, neither is "generate press" but we cross that rubicon ages ago, sadly | 12:06 |
*** jamesmcarthur has joined #openstack-tc | 13:16 | |
*** jamesmcarthur has quit IRC | 13:23 | |
*** jamesmcarthur has joined #openstack-tc | 13:26 | |
*** jamesmcarthur has quit IRC | 13:31 | |
*** jamesmcarthur has joined #openstack-tc | 13:46 | |
*** lbragstad has joined #openstack-tc | 13:54 | |
*** jamesmcarthur has quit IRC | 14:04 | |
*** jamesmcarthur has joined #openstack-tc | 14:05 | |
*** jamesmcarthur has quit IRC | 14:10 | |
smcginnis | Hmm, "runs on top of bare metal or any cloud". OpenStack is an any cloud. | 14:11 |
smcginnis | But then "should be running Kubernetes on bare metal and that there’s no need for a middle layer" | 14:11 |
smcginnis | From what I've seen, the number of folks running k8s on bare metal is a much smaller percentage than on cloud. | 14:12 |
jroll | it's almost like managing bare metal isn't worth the benefits of removing the hypervisor, for most people :) | 14:13 |
*** mriedem has joined #openstack-tc | 14:13 | |
smcginnis | There was actually a presentation at one of the events by a major bank where they described how they started out with bare metal but eventually realized they needed to run their containers in VMs due to needing isolation, hitting max open file handles, and things of that sort. | 14:15 |
* cdent had fun with max inotify watchers recently (and containers) | 14:17 | |
jroll | yeah | 14:17 |
zhipeng | This is like a never ending education task... | 14:17 |
bauzas | smcginnis: yeah but some compagnies (including mine :( ) think CNV will fix this | 14:17 |
jroll | I also can't believe the messaging that container security doesn't matter unless running third party code | 14:18 |
bauzas | well, Kube hitted its first massive CVE | 14:18 |
bauzas | I guess minds will slowly change | 14:18 |
bauzas | but whatever, the message is gone | 14:19 |
bauzas | which is a bit sad to me | 14:19 |
* jroll s/doesn't matter/is FUD/ | 14:19 | |
*** ianychoi has joined #openstack-tc | 14:20 | |
smcginnis | Yeah | 14:20 |
zaneb | jroll: yeah, because third-party code is always of the utmost quality and never gets pwned <rolls eyes> | 14:24 |
zaneb | er, first-party | 14:24 |
jroll | zaneb: indeed :) | 14:25 |
smcginnis | And why would anyone need to run third party code. Just write your own. | 14:25 |
fungi | my take on the whole "isolation is fud if you run trusted code" argument is that my parents trust the software on their laptops and it only gets hacked a few times a year | 14:27 |
bauzas | take the blue pill | 14:27 |
jroll | my take on it is that I barely trust my own code, let alone my teammate's :) | 14:28 |
smcginnis | fungi: Very true | 14:28 |
cmurphy | jroll: lol | 14:28 |
dims | bauzas : need to separate the CNCF figure heads from actual leaders in kubernetes community | 14:36 |
bauzas | dims: I'm not expert on the CNCF board but I suspect it's corporate-basis ? | 14:37 |
bauzas | if so, could we imagine some stand from the CNCF clarifying things ? | 14:37 |
bauzas | stance* | 14:37 |
dims | Dan is CNCF staff, serving at the pleasure of the CNCF board, which is a club of paying members | 14:37 |
bauzas | I don't get then why he said that | 14:38 |
bauzas | if he was running a company, then I could have understood | 14:38 |
*** jamesmcarthur has joined #openstack-tc | 14:38 | |
dims | me neither ... but we should not take his word as the sentiment of folks in kubernetes | 14:38 |
smcginnis | He's always had a very anti-OpenStack stance. | 14:39 |
bauzas | of course, like I said, I take it as an individual PoV which doesn't reflect the community standpoint | 14:39 |
smcginnis | ++ | 14:39 |
dims | +1 | 14:39 |
bauzas | but I still don't understand the need for such controversial thoughts | 14:39 |
bauzas | it's not like the CNCF needs to take a position *against* OpenStack | 14:40 |
dims | again, don't know how much of it was cherry picking by the reporter | 14:40 |
bauzas | and like dims said, that's even not a shared thought | 14:40 |
dims | right | 14:40 |
bauzas | meaning, I don't get the benefits of such standpoint | 14:40 |
bauzas | attracting more contributors ? heh | 14:41 |
gmann | dims: that's imp, any original talks link from him somewhere other than that report. | 14:41 |
dims | theres openstack people everywhere in the sigs doing good work | 14:41 |
smcginnis | So many familiar faces at kubecon. | 14:41 |
*** jamesmcarthur has quit IRC | 14:44 | |
smcginnis | tc-members: The office is open. | 15:00 |
jroll | whee. | 15:00 |
gmann | o/ | 15:00 |
aspiers | does the foundation have an IRC channel? | 15:00 |
ttx | o/ | 15:01 |
ttx | aspiers: probably the closest would be #openstack-foundation | 15:01 |
aspiers | ttx: thanks :) | 15:01 |
ttx | hmm, /me rechecks | 15:01 |
aspiers | I guess I can just email summitreg@ in this case | 15:01 |
zaneb | aspiers: there's also #openstack-board but it is pretty lightly populated | 15:02 |
ttx | aspiers: for events we have #openstack-ptg and #openstack-forum | 15:02 |
aspiers | zaneb: OK thanks, don't need the board on this occasion | 15:02 |
e0ne | hi | 15:02 |
ttx | #open-ifnra-summit has seen some traffic too | 15:02 |
aspiers | ooh, didn't know about any of these, thanks :) | 15:03 |
mnaser | add them to the list of channels you can't keep up with :D | 15:03 |
fungi | #openstack-board mostly gets used by non-board-member participants in board meetings as backchannel communication, and sits silent the rest of the time | 15:03 |
aspiers | mnaser: exactly | 15:03 |
gmann | tc-members: Gentle reminder to add your TC/cross community ideas for forum or any general topic you want to lead like retrospective/U cycle goal etc - https://etherpad.openstack.org/p/DEN-Train-TC-brainstorming | 15:03 |
mnaser | fungi: how has # of voters turned out so far? | 15:04 |
aspiers | so, if we are not close to trains this Denver summit, can the U release be called "Untrained"? | 15:04 |
ttx | aspiers: ignore me, I'm jetlagges -- most of the channels I suggested actually do not exist | 15:04 |
aspiers | ttx: lol | 15:04 |
ttx | -forum and -ptg do exist though :) | 15:04 |
fungi | the -forum and -ptg channels similarly see little activity until just before and during their respective events | 15:05 |
smcginnis | ttx is having IRC channel hallucinations. :) | 15:05 |
lbragstad | o/ | 15:05 |
ttx | Is IRC the same as a mailing-list? man, i should take vacation more often, it's refreshing | 15:05 |
dhellmann | tc-members: it would help me prep for next week if I knew who was planning to stand for chair next cycle. If you know, please contact me (privately if you wish) | 15:05 |
lbragstad | cdent appreciate the reviews on https://review.openstack.org/#/c/637025/ | 15:07 |
cdent | lbragstad: I appreciate (more!) that you wrote it | 15:08 |
fungi | mnaser: Total authorized voters: 1389; Actual votes cast thus far: 231 | 15:08 |
cdent | it's really very good | 15:08 |
fungi | so 16% | 15:08 |
fungi | or 17% | 15:08 |
bauzas | only 1389 ? I thought it was more | 15:11 |
bauzas | but okay | 15:12 |
mnaser | 1389 ATCs i guess | 15:12 |
bauzas | https://civs.cs.cornell.edu/cgi-bin/results.pl?id=E_f773fda2d0695864 showed 1636 ballots for 403 voters | 15:12 |
bauzas | anyway | 15:13 |
bauzas | thanks for the update | 15:13 |
ttx | Rocky had 384/2025 so 19% | 15:14 |
ttx | Stein was unusually well attended with 403/1636 (24%) | 15:15 |
mnaser | 300 dropped every cycle? | 15:15 |
*** e0ne has quit IRC | 15:16 | |
ttx | mnaser: on average over the last 4 years yes. We lost less lately | 15:17 |
ttx | Actually it's recruitment that is dropping. Our retention rate was increasing last time I looked | 15:18 |
*** jamesmcarthur has joined #openstack-tc | 15:18 | |
ttx | Constantly adding new projects was sustaining artificial growth | 15:18 |
*** e0ne has joined #openstack-tc | 15:19 | |
ttx | anyway, tldr: single numbers do not tell the whole story | 15:19 |
smcginnis | ATC count over the last several cycles - http://paste.openstack.org/show/746509/ | 15:19 |
mnaser | if anyone's interested, the cnf testbed sync is happening in a 10 minutes | 15:20 |
ttx | mnaser: what's that? | 15:20 |
mnaser | https://www.cncf.io/announcement/2019/02/25/cncf-launches-cloud-native-network-functions-cnf-testbed/ | 15:21 |
fungi | ttx: cncf community call about https://github.com/cncf/cnf-testbed/ | 15:21 |
ttx | mnaser: if anyone attending can call out the false dichotomy (Kubernetes OR OpenStack) I think we'll be covered :) | 15:22 |
*** jamesmcarthur has quit IRC | 15:22 | |
fungi | basically the thing the kubernetes community is trying to demo forward-thinking ideas about nfv alternatives in containers with, and show performance gains with avoidance of userland/kernelspace context switching and directly passing network payloads between namespaces, but which lf/cncf staff are trying to leverage as a competition between kubernetes and openstack | 15:23 |
smcginnis | ttx: In case you missed it, there was a conversation about a presentation that was released with that pointed out performance differences between CNFs vs VNFs. | 15:23 |
ttx | ack | 15:24 |
fungi | the cnf vs vnf performance bits were interesting. where it broke down was also positioning it as a measure of openstack vm overhead vs kubernetes running on bare metal via terraform | 15:24 |
fungi | and apples-to-oranges comparisons of openstack and kubernetes deployment timeing | 15:25 |
ttx | I feel like it's hard for OSF staff to make that false dichotomy point without sounding defensive / responding to the obvious challenge | 15:25 |
ttx | some other party that happens to be using both would be a better fit | 15:25 |
bauzas | as always, usecases decide the architecture | 15:25 |
fungi | well, they just found out that 5g is the hot new buzzword and seem to be scrambling to convince teh telecommunications industry that they should abandon openstack because kubernetes can totally do it all as long as you convince your vendors to magically rewrite all their functions to work in containers on the same kernel and also get machines big enough to put all those network functions on the same | 15:26 |
fungi | physical device | 15:26 |
*** Luzi has quit IRC | 15:27 | |
bauzas | that's a multi-year effort for kube to match with OpenStack abilities on VNFs | 15:27 |
bauzas | and 5G is at the door | 15:27 |
fungi | (because as soon as they're on different physical servers the real world sets in and introduces many orders of magnitude drops in performance which overshadow any performance gains from avoiding virtualization) | 15:27 |
fungi | and then there's the part where kubernetes still doesn't actually provision and manage your hardware for you | 15:28 |
persia | While it probably doesn't help anyone, I've found the narrative of running kubernetes on an openstack cloud ends up being fairly useful in terms of real-world deployments. At least from my perspective, continued discussion about "openstack vs. kubernetes" just makes it hard to use either (leading to some customers deciding to just use AWS because it is easier). | 15:30 |
evrardjp | fungi: I am pretty sure there are lots of tools for that | 15:30 |
fungi | yep | 15:31 |
dims | fungi - seen this? https://github.com/metalkube/ | 15:31 |
fungi | i know of one which seems to work pretty well for people running kubernetes. it's called openstack ;) | 15:31 |
dims | LOL +1 | 15:31 |
bauzas | we should stop reacting on kubernetes | 15:32 |
fungi | dims: i knew of metalkube (russellb is/was helping with that right?) | 15:32 |
bauzas | we have usecases to implement, let's focus on those | 15:32 |
dims | fungi : right | 15:32 |
dims | bauzas : +1 | 15:32 |
fungi | bauzas: i agree, we only lend credence to their specious arguments by letting them eat up our otherwise valuable time | 15:33 |
bauzas | competition is sane btw. | 15:33 |
bauzas | fungi: totally agree with you | 15:33 |
bauzas | a sane competition is OK | 15:33 |
zaneb | fungi: also dhellmann | 15:33 |
bauzas | FUD is not OK | 15:33 |
dims | zaneb : i did not want to say that out loud :) | 15:33 |
fungi | it can be someone else's job to refute those. i think the public is starting to become weary of the fake competition the lf/cncf staff seem to want to drive | 15:33 |
dims | on the + side, kops now supports deployment to openstack! if someone missed that - https://github.com/kubernetes/kops/blob/master/docs/tutorial/openstack.md | 15:34 |
zaneb | dims: first repo I clicked on the commit was Merge pull request #40 from dhellmann so I assume it's not a secret :) | 15:34 |
fungi | dims: also metalkube is openstack, in part. so... point proven i suppose | 15:35 |
dims | indeed | 15:35 |
zaneb | yes, it's standalone Ironic | 15:35 |
zaneb | so not inconsistent with what we've been saying the past couple of days in this channel | 15:36 |
fungi | indeed | 15:36 |
*** jamesmcarthur has joined #openstack-tc | 15:42 | |
*** jamesmcarthur has quit IRC | 15:46 | |
*** jamesmcarthur has joined #openstack-tc | 15:47 | |
*** dangtrinhnt has quit IRC | 16:15 | |
*** dangtrinhnt has joined #openstack-tc | 16:15 | |
*** e0ne has quit IRC | 16:16 | |
*** e0ne has joined #openstack-tc | 16:23 | |
*** dtantsur is now known as dtantsur|afk | 16:55 | |
*** e0ne has quit IRC | 17:05 | |
zaneb | I'm trying to vote but it's *really* hard to rank this group of candidates | 17:24 |
*** jamesmcarthur has quit IRC | 17:27 | |
*** jamesmcarthur has joined #openstack-tc | 17:37 | |
mriedem | if only we had integration testing of all the NFV code that was horked in since juno... | 17:45 |
mriedem | and continues to be horked in | 17:45 |
*** jamesmcarthur has quit IRC | 17:45 | |
mriedem | "hey we have a shit load of code....with no idea if it continues to work" | 17:46 |
dims | zaneb : fungi : bauzas : ttx : https://twitter.com/dims/status/1101176194651832320 | 17:46 |
dtroyer | mriedem: starlingx exercises _some_ portion of that code, right? would it be helpful to try to get specific about identifying it? | 17:48 |
*** jamesmcarthur has joined #openstack-tc | 17:48 | |
mriedem | dtroyer: L77 https://etherpad.openstack.org/p/nova-ptg-train | 17:49 |
dtroyer | hrm, I thought Intel 3CI wasn't dead, is it really? | 17:49 |
mriedem | it's been dead | 17:49 |
mriedem | mostly just bones at this point | 17:49 |
dtroyer | lemme dig in to that a bit more, the responsibility for it has moved (a few months ago) and I understood it to be running | 17:50 |
mriedem | sean-k-mooney is probably the person to talk to, | 17:51 |
mriedem | since he was managing it when he was at intel | 17:51 |
clarkb | mriedem: when you say NFV code do you mean things like pci passthrough/sriov and/or NUMA support which has typically been tested by third party CI due to hardware requirements? | 17:51 |
dtroyer | fwiw, we are setting up to test the numa live migration, both directly in Nova as well as stx's use of it | 17:51 |
mriedem | clarkb: yes | 17:51 |
mriedem | dtroyer: ok that would be good because the numa live migration series proposed for stein might not make it because of a lack of integration testing | 17:52 |
*** jamesmcarthur has quit IRC | 17:52 | |
dtroyer | right, and that is exactly what we are looking at | 17:52 |
mriedem | i'm starting to burn out (more) on piling features into nova without integration testing or long-term maintainers | 17:52 |
mriedem | we also had a forum/summit session about testing this stuff awhile back, i'm trying to find that | 17:53 |
*** jamesmcarthur has joined #openstack-tc | 17:53 | |
dtroyer | mriedem: understood, I am too and I don't even live in that world directly… trying to get the right people to care about this in a large organization is oddly hard given how important the features are in their powerpoints | 17:53 |
* cdent makes parliametary rumbilngs in support of mriedem | 17:54 | |
cdent | rhubarb! | 17:54 |
ttx | dims: nice! | 17:54 |
clarkb | mriedem: thanks, wanting to make sure there isn't a reason we can't test it upstream beyond effort but sounds like the hardware constraints remain a potential problem | 17:55 |
dtroyer | I am afraid it takes a combination of carrot and stick to get attention to this… we happen to have a convenient meeting this afternoon where I can raise it yet again… | 17:56 |
dims | ttx : spilled over to https://lists.cncf.io/g/cncf-toc/message/2938 | 18:14 |
*** cdent has left #openstack-tc | 18:19 | |
*** cdent has joined #openstack-tc | 18:19 | |
*** jamesmcarthur has quit IRC | 18:37 | |
fungi | dims: this for me underscores the difference in approach between foundations... osf basically operates on a principle of "if you can't find something nice to say about another open source project, don't say anything at all" | 18:51 |
*** e0ne has joined #openstack-tc | 18:51 | |
fungi | i'm hoping that philosophy becomes infectious | 18:51 |
dims | totally agree fungi ! | 18:52 |
fungi | and as bauzas said, we've got plenty of actual problems to concern ourselves with anyway ;) | 18:54 |
*** e0ne has quit IRC | 19:03 | |
hogepodge | I've looked a bit at their methodology, and it's a bit troubling. As with any performance metric, it varies wildly on configuration and tooling, and I'm not sure how to produce rigorous results in what appears to be a biased environment, where that bias may just be not having expertise in the tools. | 19:15 |
hogepodge | For example, the choice of chef for deployment means they picked one of the tool chains that has the least amount of support, and deploying from packages means you add fetch and install time to every node, where in the case where you deploy from containers all that time happens at the container build stage which isn't captured. | 19:16 |
hogepodge | When the test bed is framed as a competition rather than a neutral place to try and do science, it's difficult to justify allocating the resources and knowledge to participate in it. | 19:17 |
fungi | having at most skimmed their results, my biggest issue is they're using it as an opportunity to measure a variety of wholly unrelated things. how long it takes to deploy openstack or kubernetes has little if anything to do with measuring throughput between containers vs throughput between virtual machines | 19:18 |
hogepodge | As someone who's really tried to build ties between the communities it's damaging and, from my personal experience, doesn't capture the feelings of the people I've worked directly with. | 19:18 |
fungi | each is a potentially interesting comparison in its own right, but a proper case study would have focused on one of those, not both | 19:18 |
*** whoami-rajat has quit IRC | 19:20 | |
fungi | as for openstack-chef having the least amount of support of our available deployment methods, i can't really speak to that. scas probably has opinions there and if they do want to stick with chef maybe he has insights on how to make that option more efficient for them too | 19:20 |
hogepodge | even setting that aside, a chef or puppet install will do the software installation at the same time as configuration, a container approach does the install as a pre-step, so that time isn't captured in benchmarks | 19:23 |
hogepodge | but there may be a real opportunity to investigate the performance of our network stack in a productive way, but it's hard to take advantage of those sorts of resources and opportunities if it's presented in a negative and potentially biased light. | 19:25 |
fungi | right, from that perspective they at least ought to compare similar deployment methodologies. if they're going to deploy openstack from source or packages then they should be deploying kubernetes the same way. if they're going to count server provisioning into the total time for one then they should count it on the other as well, and so on | 19:29 |
*** jamesmcarthur has joined #openstack-tc | 19:56 | |
*** e0ne has joined #openstack-tc | 20:16 | |
ttx | https://twitter.com/OpenSourceOrg/status/1101202880239886336 | 20:46 |
cdent | huzzah | 20:47 |
cmurphy | nice | 20:47 |
*** cdent has quit IRC | 20:48 | |
smcginnis | Excellent | 20:49 |
zaneb | yay | 20:58 |
*** e0ne has quit IRC | 21:19 | |
*** e0ne has joined #openstack-tc | 21:42 | |
*** e0ne has quit IRC | 21:44 | |
*** e0ne has joined #openstack-tc | 21:45 | |
*** e0ne has quit IRC | 22:01 | |
tonyb[m] | I'm hearing rumors that rackspace has just made the entire RPC (rackspace private cloud) team redundant. This means that some of those great members of our community will be going thru lot right now. Can we look at ways to help them? Perhaps by extending the TSP? to get them to Denver | 22:24 |
tonyb[m] | I know that helped me a lot around the Boston stoney | 22:25 |
tonyb[m] | Summit sorry | 22:25 |
tonyb[m] | ... I also know that's a foundation level decision but many foundation staff are here :-P | 22:26 |
smcginnis | Yeah, sounds like there were quite a few hit by this. | 22:26 |
clarkb | aiui the application was reopened until at least noon tomorrow | 22:27 |
tonyb[m] | Ahh cool | 22:29 |
* mnaser has personally reached out and invited people to reach out to me personally as we're hiring | 22:29 | |
tonyb[m] | Bigger than OSIC if I understand correctly | 22:29 |
mnaser | this hurts OSA a lot | 22:29 |
tonyb[m] | mnaser: thanks! | 22:29 |
clarkb | tonyb[m]: the constraint there being reviews have to happen as well as being able to book the travel details after. | 22:29 |
clarkb | and noon tomorrow is middle america time I think | 22:30 |
tonyb[m] | clarkb: all good anything helps | 22:31 |
clarkb | TheJulia: jroll (but TC at large too) we've just been asked on the discuss list if we would give up https://pypi.org/project/nimble/#description so that someone else can use the name. THe description says this is realted to ironic so you amy know what that would entail on our end | 22:32 |
mnaser | yeah, i saw that. | 22:33 |
mnaser | i was also wondering if there's any code at all that's being shared, because no one came and said "here's our project that we want to use the namespace for" | 22:34 |
smcginnis | IIRC, that was a project that got renamed. | 22:34 |
mnaser | i feel like that'd be the .. easier thing to do | 22:34 |
smcginnis | Because "nimble" is the name of a company. That also happens (or happened) to be involved in Cinder as a storage company. | 22:34 |
fungi | clarkb: TheJulia: jroll: that is the project now known as "mogan" | 22:34 |
clarkb | mnaser: not sure I understand? | 22:34 |
fungi | based on the research i just completed | 22:34 |
clarkb | ah ok so we already renamed it to somethign else | 22:35 |
smcginnis | fungi: Oh right. | 22:35 |
*** prometheanfire has joined #openstack-tc | 22:35 | |
fungi | at the beginning of 2017 | 22:35 |
smcginnis | Which is also dead, if memory serves. | 22:35 |
smcginnis | Bottom line, I think it's fine to free that up on pypi. | 22:35 |
fungi | there was a project-config patch to rename it to mogan over the end of 2016 holidays | 22:35 |
TheJulia | I would concur that we should be able to release it, but I'm not sure ironic folks are the right people to be the explicitly bless it. | 22:36 |
tonyb[m] | Who does pypi trove say owns it? | 22:37 |
tonyb[m] | Apart from openstackci :-P | 22:38 |
clarkb | https://pypi.org/user/Zhenguo/ | 22:38 |
smcginnis | Someone from Huawei. We renamed it and stopped working on it, so I guess you can take me with my Huawei hat on saying OK unless you want to track him down. | 22:38 |
clarkb | I think if we can show we renamed nimble to mogan then we should be fine to release it assuming the TC wants to release it (whcih also assumes this was an official project) | 22:39 |
smcginnis | Not official. | 22:39 |
smcginnis | They pitched it in Sydney but there were too many unanswered questions and issues with Nova API usage. | 22:40 |
smcginnis | Abandoned patch to make it official - https://review.openstack.org/#/c/508400/ | 22:41 |
smcginnis | But anyway, definitely renamed from Nimble, so no reason to squat on that name. | 22:41 |
clarkb | ya so I think we can go ahead and hand it over. Does anyone know if there is an official process for that or we just add their user and they remove ours? | 22:42 |
jroll | I agree with everything smcginnis said | 22:42 |
smcginnis | clarkb: Normally I would say "ask infra", but... :) | 22:42 |
smcginnis | That does sound right though. | 22:42 |
smcginnis | For historical trivia, here's the etherpad from the Sydney Forum - https://etherpad.openstack.org/p/SYD-forum-baremetal-ironic-mogan-nova | 22:43 |
clarkb | smcginnis: ya I've never done one before. I think tempest did though? mtreinish may know | 22:43 |
smcginnis | clarkb: They need to be able to update it and publish, so I think as long as they are added as an owner they should be able to take it over and do whatever they want with it. | 22:44 |
mtreinish | clarkb: when I did it for tempest I reached out to Richard Jones directly, who in turn asked the old owner of tempest for me. | 22:44 |
mtreinish | they just added my pypi account to the owners list on the package | 22:44 |
mtreinish | and I then backed up everything locally in a few places, cleaned up the metadata, and uploaded openstack tempest | 22:45 |
clarkb | mtreinish: cool so adding the other individual's account is basically all we have to do on our end | 22:45 |
mtreinish | yeah, or at least that's all that happened when I took over tempest | 22:45 |
mtreinish | heh, actually this made me wonder if I can even get the old non-openstack tempest tarballs I have locally to run anymore | 22:49 |
mtreinish | the last pypi release was from 10 years ago (although they had a tarball on google code from 9 years ago, which was never pushed to pypi) | 22:50 |
*** jamesmcarthur has quit IRC | 23:46 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!